Automated machine-based classifiers can be used to identify features of interest in a sample based on one or more sample images. Classifier training can be a time-consuming process and can be prone to operator error.